3535b released

**Eugene Gavrilov** · Sep 27, 2003, 06:38 PM

hi there

3535b is out [development release, experimental]
========================================

[changes]:
Added support for 24/96 playback and recording for Audigy2-based boards (EXPERIMENTAL)

Improved power management support
Improved hardware compatibility with certain audio cards
Fixed bug with AC97 contoller power management
Improved SMP / HyperThreading support (EXPERIMENTAL)

Improved real-time panning quality
Added option to choose Wave panning method (true pan or balance) - by command-line tool

Added support for horizontal peak meter
Added a shortcut for the peak meter to the kX Mixer page
Improved IO Status window
Improved 'mute speakers when headphones are plugged-in' option

Added option to choose the desired speaker spatialization method (by using command-line tool)

kX DSP window enlarged
Added real-time update option for kX Editor and kX API (EXPERIMENTAL)
Updated Filter effects (may be not 100%-compatible with previous)
Added APS Compressor + with Side Chain input
Added MX6 (custom mixer) DSP effect (described here: http://www.kxproject.com/mx6_help.php
Added high-quality headphones positioning effect (HpSp
Updated XRouting effect to support 8-point spatialization
Prolog and Epilog effects re-written to support amplification (EXPERIMENTAL
Improved kX Mixer indication for inputs & outputs (levels measured in dB
(NOTE: if you have any pre-3535 setting files, the new prolog/epilog
features won't be available until you upload new prolog/epilog effects
into the DSP. Resetting settings will upload the new effects automatically
The SDK License Agreement has been updated<br>

Danish translation added
Polish translation updated
Improved System Tray icon

===============
TODO:
fix kX Editor's 'Update' function to restore 'output' connections, too
finally check SMP/HyperThreading and AC97 hibernation issues
make dB sliders smoother (?)
===============

any feedback welcome
please read the changelog carefully!

URL: http://kx.maincore.nl/beta/kxdrv3535b-update.exe
the SDK: http://kx.maincore.nl/beta/kxdrv3535b-sdk.exe

users who have 3535a installed: please reset settings after upgrade

/Eugene

Robert McClelland · Sep 27, 2003, 11:13 PM

I be downloading to test out now bud let you now thanks

Robert McClelland · Sep 28, 2003, 12:03 AM

I foung one bugs and no headphone plug in at all only FXMix 2 shows up in DSP Want me take snap of it for you bud send by email? Or you want me installed the headphone plugin?
Also on FX 1 and 2 one makes front speakers louder right? and FX2 makes back speakers louder right bud?

kode54 · Sep 28, 2003, 12:38 AM

Doppler still seems to be backwards, unless FMOD's 3D sample is using inverted listener velocity, or FMOD itself is doing something funny. Hmm, doesn't happen with software 3D...

zystudio338 · Sep 28, 2003, 04:00 AM

I Install 3535b, use it with SONAR 2.2 record a Violin solo,about 10-20 minutes sonar and winXP is dead:Repeat endlessly a note.
I using sonar in 'kernel streaming' mode,Audigy 1 ,I Have to reintall 3534 , everything is OK!

**Eugene Gavrilov** · Sep 28, 2003, 09:16 AM

>> no headphone plug

it is not present in the default DSP layout
you need to upload it manually

have a look here: http://kx.maincore.nl/headphones.gif (a DSP snapshot)
(settings file (audigy2 ex-compatible): http://kx.maincore.nl/headphones.kx

note: in this configuration the 'front & rear' option is active; the headphones are connected to the 'front' output, while the speakers are connected to the 'rear' one. headphones level is reduced (you may need to restore it to its default)

>> Doppler still seems to be backwards, unless FMOD's 3D sample is using inverted listener velocity

what is the exact sample that causes the problem? (I have FMOD somewhere... I will check this issue then)

>> winXP is dead

do you have an SMP/HT machine?

====

has anybody tested the ac97 hibernation bug?

/Eugene

Eminos · Sep 28, 2003, 12:04 PM

Quote:

has anybody tested the ac97 hibernation bug?

Yes...It seemes to work now with the 3535b ...

/Eminos

Robert McClelland · Sep 28, 2003, 12:40 PM

Only bug I detect is there volume like contral on the 3-D part in the router? Everthing else is awesome

nice and loud hehe

thanks for that Trimbe stuff

Robert McClelland · Sep 28, 2003, 04:00 PM

I got good news everone I gotton the center speaker working on the 5.1 setup

Which effect is the thump for the sub woofer during the speaker test?

kode54 · Sep 28, 2003, 07:39 PM

>> what is the exact sample that causes the problem? (I have FMOD somewhere... I will check this issue then)

Samples\3D\3D.exe, featuring two static 3D sounds and a moving listener.

ookzdvd · Sep 29, 2003, 09:05 PM

I think the volume control over 5.1 (ac3) playback is still broken,
I still can hear the sound even I've already slide down the volume to 0.

Thank you.

kode54 · Sep 29, 2003, 09:13 PM

AC3 bypass mode?

amol0010 · Sep 29, 2003, 11:42 PM

Eugene, i would like you to have fix the bass bug in the rear and headphone output mode, this is driving me insane,
And even with the phase DSP registered and connected, it doesnt help, the bass still distorts and lacks any punch at all. The front output delivers much heavier and tighter bass than the rear and the headphone output mode. isnt there a way to get the exact same sound on the rear and heaphone mode output ( the center/subwoofer -> headphone mode), in terms of bass quality.

Thanks for any reply given at all.

**Eugene Gavrilov** · Sep 30, 2003, 04:48 PM

>> I think the volume control over 5.1 (ac3) playback is still broken

I've investigated this problem

* when in 'ac-3 decoding mode' (software decoding by the driver), the 'volume' control is unavailable since microsoft kernel-level driver doesn't use the volume control for spdif/encoded outputs [microsoft limitation]
* when in 'ac-3 bypass mode' the volume control is theoretically impossible [since the driver doesn't modify the encoded data]
* when in 'ac-3 multichannel mode' (the player decodes the data, not the driver): the player is responsible for changing the volume for =all= the channels. according to my tests, most players change only front L/R channels
this seems to be a bug in these players... (although, Creative drivers might use a kind of a workaround, but they break the specification then...)

I could probably add an additional option to follow CL-like volume control... ['compatibility option']

====

>> the bass still distorts and lacks any punch at all

please start a separate topic with this question (or find your previous one)
post your hardware information
try to figure out your ac97 / i2s codecs (located on-board)
-- it seems some of your codecs aren't fully initialized...

/E

ookzdvd · Sep 30, 2003, 08:56 PM

Quote:

Originally posted by Eugene Gavrilov
>> I think the volume control over 5.1 (ac3) playback is still broken

I've investigated this problem

* when in 'ac-3 decoding mode' (software decoding by the driver), the 'volume' control is unavailable since microsoft kernel-level driver doesn't use the volume control for spdif/encoded outputs [microsoft limitation]
* when in 'ac-3 bypass mode' the volume control is theoretically impossible [since the driver doesn't modify the encoded data]
* when in 'ac-3 multichannel mode' (the player decodes the data, not the driver): the player is responsible for changing the volume for =all= the channels. according to my tests, most players change only front L/R channels
this seems to be a bug in these players... (although, Creative drivers might use a kind of a workaround, but they break the specification then...)

I could probably add an additional option to follow CL-like volume control... ['compatibility option']

====

/E

I think my problem is the first one, 'ac-3 decoding mode'

You are right the volume control only change the L/R channel only.
not all the channels.

Thank you for your explaination and investigation

ArconX · Oct 2, 2003, 08:56 AM

I'm sorry to say but the 3535a/b release are not worth the name.

3535b has even become worse! then 3535a.

Example:

3535a was playable with winamp, but locks game's & your pc that tried to run it.

Now with 3535b:

Trying to Reset Settings: Lockup the entire pc.
Playing Winamp3: Errorbox of a missing file.
Retry playing Winamp3: Lockup the entire pc.
Playing Games, the same deal.

3534 works perfectly (and stable) on the same system, the only change that may affect the same system that i see is the Hypertrait changes. All this seem to be related to the "new" & improved SMT/Hypertrait support. Note: I'm running a 3.06 HyperTrait P4.

Eugene, are you running a PC with hypertrait? Becouse i from what i see, i don't think so. My advice is to remove that hypertrait code until you can personaly test it on a system becouse now it's just something to lock systems (and personaly, i don't see how a improved hypertrait support will help the drivers, sinds all hypertrait cpu's are 2.4Ghz+, and you don't see the impact of the drivers on those speeds) ...

My system specs are posted in the old 3535a trait in case you need them.

**Eugene Gavrilov** · Oct 2, 2003, 04:36 PM

>> 3535b has even become worse! then 3535a

that's why they are 'development releases'

>> are you running a PC with hypertrait

no, of course

>> until you can personaly test it on a system

I doubt this will happen soon.. probably, in 2-3 years

>> improved hypertrait support will help the drivers

some people have reported problems with HT enabled (spinlock issues) -- I simply tried to fix them

of course, the next 3535c release will contain the original interrupt handling code

/E

calinb · Oct 3, 2003, 05:40 PM

Quote:

Originally posted by Eugene Gavrilov
I could probably add an additional option to follow CL-like volume control... ['compatibility option']

Eugene, Yes--that would be a very useful option for those of us with KX-powered HTPCs! (Generic remote controls that can be programmed to control player ap volume would control all channels again!) I asked Gabest if he could fix it in MPC, but he didn't seem very hopeful.

Can I bribe you?--I just donated to KX

**Eugene Gavrilov** · Oct 4, 2003, 06:45 PM

>> that would be a very useful option for those of us with KX-powered HTPCs

ok, I will add it to my 'todo' list

but I will need to check if it breaks any audio applications...

/E

kode54 · Oct 4, 2003, 06:58 PM

You say that front left/right problem with DVD players is a problem with the players. I have found a similar problem with a crossfading output for Foobar2000. It uses IDirectSoundBuffer::SetVolume() to fade streams out on 10ms intervals before closing them. The effect only seems to work on the front left/right channels. No such problem if I force the output to disable hardware mixing, but then the rear channels are barely audible. (I can hear them with my headphones if I max the volume levels in the mixer, from their usual levels of 6 and 8, and I can hear the fading.)